IMU Attitude Estimation for Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Vehicular navigation systems face challenges in accurately determining vehicle position and velocity when inertial measurement units (IMUs) are not rigidly mounted and undergo changes in relative attitude during undocking and re-docking, especially due to vehicle acceleration and motion.
Innovation Solution
The system estimates the relative attitude between the IMU and the vehicle frame of reference using pitch, roll, and yaw angles, and iteratively minimizes estimation errors by correlating acceleration and angular velocity measurements, allowing for accurate translation of IMU measurements to the vehicle frame without requiring a rest state.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the IMU is rigidly mounted on the vehicle with fixed reference axes aligned to vehicle geometry, then the position and velocity determination is simple and accurate, but the system cannot accommodate undocking and re-docking operations and loses adaptability
Solution Approach 1:
The system transitions from a static rigid mounting assumption to a dynamic attitude estimation model that continuously adapts to changes in IMU orientation. The relative attitude between the IMU and vehicle frame is estimated in real-time using pitch, roll, and yaw angles, allowing the system to accommodate undocking and re-docking operations while maintaining navigation accuracy.
Solution Approach 2:
The system introduces additional parameters (pitch, roll, yaw angles) to describe the relative orientation between the IMU and vehicle frame. By estimating these attitude parameters dynamically, the system can transform IMU measurements into the vehicle frame of reference even when the IMU is not rigidly mounted, resolving the contradiction between adaptability and complexity.
2Productivity
If the relative attitude is estimated using projection of gravity vector when vehicle is at rest, then the estimation is simple, but the system cannot determine attitude during vehicle motion and loses continuity
Solution Approach 1:
The system implements continuous attitude estimation by combining gravity vector projection (when available) with iterative optimization using accelerometer and gyroscope measurements (when vehicle is in motion). This ensures uninterrupted attitude determination throughout all vehicle states, maintaining productivity while preserving precision through multi-source fusion.
Solution Approach 2:
The system uses iterative optimization that incorporates feedback from both accelerometer measurements (gravity projection) and gyroscope measurements (angular velocity integration). This feedback loop continuously refines the attitude estimation, resolving the contradiction between simplicity and precision by combining multiple measurement sources in a unified estimation framework.
3Measurement precision
If the estimation error in relative attitude is minimized by iterations, then the navigation accuracy is improved, but the computational time and processing complexity increase
Solution Approach 1:
The system performs preliminary attitude estimation using gravity vector projection when the vehicle is at rest, establishing an initial accurate attitude state before motion begins. This preliminary action reduces the computational burden during subsequent motion phases, as the iterative optimization starts from a better initial condition, thereby reducing the time required to achieve convergence.
Solution Approach 2:
The system applies iterative optimization selectively based on vehicle motion state. During constant velocity motion, the system uses partial iterations or simplified estimation methods, while reserving full iterative optimization for transitions or high-precision requirements. This partial action approach balances computational time with measurement precision needs.
4Measurement precision
If the yaw angle is estimated by de-tilting the measurement after determining drive axis, then the drive direction is accurately determined, but the 0-π ambiguity remains and requires additional correlation with GNSS acceleration
Solution Approach 1:
The system introduces GNSS acceleration data as an intermediary to resolve the 0-π ambiguity in yaw angle estimation. By correlating the IMU-derived acceleration with GNSS acceleration, the system can disambiguate the drive direction while maintaining computational efficiency. This intermediary approach resolves the ambiguity without requiring complex additional sensors or processing.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate determination of vehicle position and velocity even during changes in the IMU's spatial orientation, reducing residual errors and resolving ambiguity in drive direction estimation, thus improving navigation accuracy across various vehicle motions.
Implementation Method 1
The INS uses inertial sensors to determine the vehicle's position and acceleration. Often, three dimensional inertial sensors (part of INS) comprising accelerometers and gyroscopes are deployed on the vehicle.
Implementation Method 2
projection of the gravity vector on the IMU reference axes are used to determine the relative attitude when the vehicle is at rest.
Implementation Method 3
an acceleration derived from a GNSS is correlated with the acceleration from IMU to determine the drive direction.
Data Source
AI summary
According to an aspect of the present disclosure, the relative attitude between an inertial measurement unit (IMU), present on a mobile device, and the frame of reference of the vehicle carrying mobile device is estimated. The estimated relative attitude is used to translate the IMU measurement to the vehicle frame of reference to determine the velocity and position of the vehicle. As a result, the vehicle position and velocity are determined accurately in the event of undocking and re-docking of the mobile device from a docking system in the vehicle. The relative attitude is estimated in terms of pitch, roll, and yaw angles.